Ripple published a Boston Consulting Group research report on institutional digital asset custody on October 21, arguing that banks and investment firms no longer treat asset security as a differentiator. It's the floor. What they're weighing now is governance, interoperability and whether a custody stack plugs into settlement, payments and treasury systems. The report says more than 30% of digital assets held outside centralized exchanges sit in institutional custody models — an estimate from commissioned research, not a wallet-by-wallet count.
What BCG actually measured
The 30% figure is the headline number, and it comes with a caveat attached: it's a modeled estimate, not an observed tally of every address in the market. That distinction matters because custody share is one of those metrics that gets repeated until it sounds like a census. BCG's framing is broader. Security is described as a baseline requirement — the thing you have to get right before anyone bothers evaluating you on anything else.
From there, the report lists policy controls, interoperability, and the ability to connect custody with settlement, payments and treasury as the criteria that separate shortlisted providers from the rest.
The integration problem for stablecoins and tokenized securities
This is where the argument gets concrete. For stablecoins and tokenized securities, a custodian that holds assets but can't clear approval procedures, generate the right reporting, or talk to existing back-office systems leaves institutions unable to deploy the assets at all. Holding is easy. Moving is the job.
That's why the report pushes integration so hard. A custody arrangement that forces a bank to bolt on separate reporting and approval tooling has technically solved safekeeping while quietly breaking the reason the bank showed up. The asset sits there. Nothing else works.
One custodian or several behind an orchestration layer
BCG doesn't settle on a single model. Some firms, per the report, will spread assets across multiple custodians and run a shared orchestration layer on top. Others will want one integrated provider and fewer moving parts. Both are live options, and the choice likely tracks how much internal operational complexity a firm is willing to absorb.
Ripple sells custody infrastructure, so its decision to commission and promote this report is commercially interested. Read the framing with that in mind. The underlying point — that banks want conventional institutional controls without giving up the settlement advantages that pulled them toward blockchain in the first place — is the argument Ripple has an obvious stake in winning.
What the report leaves open
The unresolved question is what a shared orchestration layer looks like in practice, and who operates it. The report describes the model without naming a standard or a provider. Until that consolidates, custody buyers are effectively being asked to pick a side between integration and flexibility, with the compatibility risk sitting squarely on their own systems teams.
